training flops
- AdaSTaR: Adaptive Data Sampling for Training Self-Taught Reasoners
- CoIDO: Efficient Data Selection for Visual Instruction Tuning via Coupled Importance-Diversity Optimization
- Faster Video Diffusion with Trainable Sparse Attention
- FlexOLMo: Open Language Models for Flexible Data Use
- Mixture-of-Recursions: Learning Dynamic Recursive Depths for Adaptive Token-Level Computation